this article has been accepted for BSOAS and will appear in the journal subsequent to editing.

The copyright belongs to Cambridge University Press.

Footnotes on a Parthian Sound Change

Agnes Korn Empirische Sprachwissenschaft, Universitt Frankfurt am Main, PF 11 19 32, 60054 Frankfurt, Germany a.korn@em.uni-frankfurt.de To the memory of Jochem Schindler (1944-1994)

Abstract The treatment of Proto-Iranian *w (PIE *t) is one of the isoglosses distinguishing Middle Persian from Parthian and thus important for Western Iranian dialectology. Its re-discussion of the Parthian development of this consonant cluster by Nicholas Sims-Williams presents a welcome occasion for some notes on the matter.1 I will argue that there is some additional evidence in favour of his suggestion that the Parthian result is not -f- as previously assumed, but a consonant cluster. I will also suggest a modification of the steps that the development takes. The Middle Persian development of *w as well as some related issues of historical phonology and Pth. orthography and Western Ir. are likewise discussed. Keywords: Parthian, Middle Persian, Western Iranian, historical phonology, isoglosses, Iranian dialectology

1. According to the classical treatments by Tedesco (1921: 199f.) and Henning (1958: 96f.), Proto-Ir. *w (PIE *t) gives Middle Persian h (MP ahr four vs. Avestan ar, OInd. catvra; MP ihil forty vs. Avestan aarsat-, OInd. catvrit-; abstract suffix MP -h < *-iya-wa-), but f in Parthian, as in afr four, afrast forty, and the abstract suffix -f in inscriptional Parthian. This interpretation of the Parthian data needs to rely on explaining the additional -t seen in -ft, the variant of the abstract suffix seen in Manichean Parthian, as an additional suffix (Tedesco 1921: 200 suggests a derivation from *-iya-wa-t-). However, this approach does not offer an explanation for the Pth. verb (present stem) <nydfr-> nifr- / (past stem) <nydfwrd> nifurd to hurry (itr. and tr.), which is likely to derive from *ni-wraya- (cf. OInd. tvar), and the noun <nydfr> nifr haste (Henning 1958: 97 n. 2).2 While

The present article is a revised version of a German paper written in 2004/05 for a collective volume that has not been published. For reasons of typographical simplicity, , w and y are used instead of , and insofar as Proto-Iranian is concerned. As per Iranological tradition, italics represent the transcription (phonemical form) for Parthian and Middle Persian, but the transliteration (graphical form) for Sogdian. Manichean Middle Persian and Parthian are quoted from and in the form of DMD unless otherwise noted. Abbreviations: Av. = Avestan; B = Buddhist Sogdian; C = Christian Sogdian; Ir. = Iranian; Manich. = Manichean; M = (Sogdian in) Manich. script; MP = Middle Persian; OInd. = Old Indic (Vedic and Sanskrit); OP = Old Persian; PIE = Proto-Indo-European; Pth. = Parthian; S = (Sogdian in) Sogdian script. For bibliographical abbreviations, see the references at the end of the article. I am indebted to Desmond Durkin-Meisterernst, Jost Gippert, Thomas Jgel, Nicholas Sims-Williams and Yutaka Yoshida for comments and discussion, and to several colleagues for their contributions acknowledged in the notes. 2 If the root had the shape PIE *terH (as sometimes assumed), the past stem would be *tH-to- > *warta-, in which case Pth. <nydfwrd> could be read niford (thus DMD 252b for the derivative <nydfwrdg> hurried).

Hennings etymology is certainly convincing, his further suggestions are less so: he assumes that the word-internal result of Proto-Ir. *w is Pth. -f- while f in ni-fr- would show the result in word-initial position (for which there is otherwise no example), and that f would have been adopted from the (unattested) simplex *fr-. This scenario is improbable not only because it implies the unlikely assumption that a cluster that is reduced to -f- in word-internal position would be retained word-initially, but also because the parallel consonant cluster PIE *d > Proto-Ir. *w is reduced to Pth. b- word-initially (Sims-Williams 2004: 540). Sims-Williams (2004: 540, 545) thus suggests the alternative solution that f is the regular result of *w in word-internal position. For afr four he assumes a dissimilatory loss of the dental elements of the consonant cluster (*[tafr] > [tafr]), a development that also occurred in this word in other Ir. languages (e.g. Bactrian four vs. regular < *f in - attain < *wana-, abstract suffix -(), Sims-Williams 2004: 542). For the word-final position, he posits a dialectal difference in the further development of *-f > -f for inscriptional vs. -ft for Manichean Parthian (Sims-Williams 2004: 543, 546). 2. This set of changes is so far based on one example each. But there seems to be additional evidence confirming Sims-Williams assumption that *w gives Manich. Pth. -ft, also implying that the abstract suffix -ft does not contain an additional suffix. 2.1 There is a word <pwrt> bridge in the Pth. hymn cycle Angad Rnn VI 57b.3 Although this is a hapax legomenon, its reading and meaning are reasonably clear. The existence of such a word in North-West Iranian is also confirmed by Gilaki purt, purd and Zazaki prd bridge.4 Etymologically it is obviously related to Avestan prtu- (cf. Boyce 1954: 194: < *prtu-). However, a derivation from Proto-Ir. *ptu- would raise a problem on the phonological side: Proto-Ir. *t following a labial otherwise and expectedly5 gives Pth. <wrd> -urd, e.g. <bwrd> burd < *bta- (past stem of <br-> carry), <mwrd> murd < *mta- (past stem of <myr-> die). Proto-Ir. *ptu- should thus have given <pwrd> purd. So it is worth considering whether Pth. <pwrt> could derive from the oblique stem *pw-, i.e. from the form that has always been seen as underlying the MP cognate puhl (*pw- > *purh > puhl, Hbschmann 1895: 195, 207, Hoffmann 1986: 171, 181 n. 20).6 The application of the change suggested by Sims-Williams for Manich. Parthian (see Section 1.) yields *pw- > *purf > *purft. Since a consonant cluster -rft is not permitted by Pth. phonotactics,7 *purft could have been reduced to purt by a dissimilation vs. the initial p- that is not unlike that in afr.
However, there are good arguments against the laryngeal (EWAia I: 685, de Vaan 2003: 56, LIV p. 655), so nifurd < *-wta- < *-t-to- seems preferable (thus e.g. Ghilain 1939: 74, Boyce 1977: 64). Weber (1994: 111 n. 11) interprets <nydfr-> asa compound related to MP dwr- run, move (according to Weber an Avestan borrowing), but MP dwr- differs from the Pth. <nydfwrd> verb in its past stems (MP dwrist and dwrd). Webers etymology also involves the problem that word-internal *dw gives Pth. <db> v (Sims-Williams 2004: 540). 3 Cf. the edition Boyce (1954: 148). DMD 287a reads /purt/, /purd/?, Boyce 1977 does not note the word. 4 None of the contemporary varieties is a direct decendant of Parthian, but they can hint at the existence of otherwise unattested words and word forms in Middle West Iranian. 5 * > Pth. ur is the regular development in labial context (Rastorgueva / Molanova 1981: 172). For Old Ir. *-t > Pth. d see Section 5. 6 The word is likely to have had an amphidynamic paradigm PIE *prtu- / *pt-- (Hoffmann 1986: 171). New Persian (classical) pul cannot come from *ptu- since this would have given purd here as well. 7 There are no Pth. tautosyllabic clusters of three consonants (DMG 3.1.1.2.3); in order to avoid them, * in old sequences of *ft does not yield *r, but probably gives *r from the outset, cf. <gryft, grypt> ()grift < *gfta- (past stem of <gyrw-> seize).

2.2 A derivation of <pwrt> purt from *purft < *pw- suggests a parallel explanation for Pth. <mwrt> murt death8 from *murft < *murf < *mw- (Nicholas Sims-Williams, p.c.). A dissimilatory loss of f in *murft is surely as motivated as it is in *purft. On the other hand, *mw- would be the oblique stem of a so far unknown Ir. stem *mtu- besides the otherwise attested *myu- (Avestan mriiu-, Old Persian (uv-) mriyu-,9 OInd. mty-), but a stem *mtu- / *mw- death is indeed reflected in Sogdian mwrw /mur/.10 This is likely to derive from the nominative and accusative forms *mu and *mum11 while a derivation from *myu- should effect a palatalisation of the vowel (Sims-Williams, p.c.). Similarly, the derivation of Pth. <mwrt> from Proto-Ir. *mti- suggested by Henning (1937: 85) should probably give <myrd>, cf. *kta- > <kyrd> (past stem of <kr-> do), *mya- > <myr-> die.12 So far as the existence of *mw- in Sogdian is concerned, the word is found in B pymwrw13 after death and in the phrase M z mwrw birth-death, B z t ( t) mwrw birth (and) death (i.e. circle of reincarnation, sasra), where z - shows a change of Old Ir. *-t that otherwise does not occur in Sogdian. So Benveniste (1940: 216) assumes a Pth. origin while the original Sogdian phrase would be zy myry birth-death.14 Indeed, Pth. <zdmwrd> zdmurd is quite well attested, and Pth. influence in the Sogdian Buddhist lexicon has been noted for other words as well.15 However, while Pth. influence in the use of Sogdian ()mwrw and in the formation of z mwrw is possible, the assumption of a direct borrowing is faced with the difficulty that the attested Pth. forms are in fact <mwrt> murt and <zdmwrd> zdmurd.16 It would also be unlikely to assume that Sogdian borrowed mwrw from the stage of Pth. *murf, since one would expect Pth. f to be rendered by Sogdian <>. Such an output may be seen in Sogd. pwtys Bodhisattva (besides variants such as pwt/yst), which could owe its <> to Pth. <bwd(y)sdf> bdisaf (thus SimsWilliams 2004: 544f., see also Section 3.).17 If mwrw is thus an inherited Sogdian form, Sogdian would show several words for death (cf. mrc, B mwrtk(y)). The stem *mu- underlying Sogdian mwrw would derive from a paradigm *mtu- / *mw-, the oblique stem of which yields Pth. <mwrt>.18

Found in Angad Rnn VII 4a (cf. Boyce 1954: 154) in several copies. On this word, see Gippert 2001. 10 In the alphabets used for the Manich. (M) and Buddhist (B) Sogdian texts, <> is used for und while the script of the Christian texts (C) has an extra letter <> for . 11 *mu- with generalised (from the oblique stem *w) is parallel to OP gu- from a paradigm *gtu- / gw- (cf. note 24). 12 <myr-> shows that the palatalising effect of a following *y is stronger than the labialising effect of m-. 13 Two attestations in Benveniste 1940 (for the attestation 8, 52 in Benveniste 1940: 269 and Gharib 1995: 337a read 8, 72) and one in the British Library Frag. 6 line 5 (rather fragmentary context), where Sims-Williams (1976: 49) reads p[y]()m(wr), but there seems to be a final -w also in this attestation (p[y]()m(wr)(w)), cf. the photo of Or. 8212/82 on the webpage of the International Dunhuang Project (http//:idp.bl.uk). 14 zy myry birth-death is found only in Benveniste (1940: 56, line 1194). Gershevitch (1946: 148) considers zmwrw as a loan translation. Gharib (1995: 453b) follows this view, which might be the reason for her reading Sogd. zmwrd (sic) /zdmurd/ and pymwr(w) /pimur/ (the paragraphs referred to in Gershevitch 1954 only note ()mwrw, though). On the other hand, she reads mwrw /muru/ (Gharib 1995: 221a). 15 Cf. Sims-Williams (1983: 139, 2004: 544), Sundermann 1982. 16 The -d of the latter against the -t in <mwrt> can be explained by association (not only by popular etymology) to the past stem murd, perhaps additionally motivated by the final of the first member of the compound. 17 However, Yoshida (2008: 344-353), providing a list of variants and attestations, argues against Parthian influence in the Sogdian word for Bodhisattva. 18 In Parthian, other terms in this semantic field include <w> death, <zgm> izm flight, exit (of the soul from the body) and Ind. loanwords found in Buddhist context (<mrn> maran, <prnybrn> parnirn, cf.
9

3. Another item to be considered in the discussion of the Pth. result of Proto-Ir. *w is the inscriptional Pth. form <nytprywt> hurried corresponding to Manich. Pth. <nydfr> etc.19 It seems that the most straightforward interpretation of the <-tp-> is tf. In this case, one might consider a modification of the changes noted in Section 1.: Proto-Ir. *w could have yielded Pth. tf first, which would be shown by inscriptional <nytpr-> nitfr-. In four, a dissimilation *[tatfr] > [tafr] <cfr> would have taken place.20 Word-final *-tf would have undergone a metathesis to -ft in Manich. Parthian and a reduction to -f in the dialect of the Pth. inscriptions, thence the abstract suffix Manich. -ft <-yft>, inscriptional -f <-py>. The output of Proto-Ir. *pw- and *mw- would have been reduced to <pwrt> purt and <mwrt> murt by the phonotactic ban on tautosyllabic clusters of three consonants (cf. note 7), either already at the stage of *purtf and *murtf or in the metathised stage of *purft and *murft. The next stage would assimilate the tf to f. This would have concerned word-internal cases of tf other than four, thence Manich. <nydfr-> nifr- and derivatives vs. inscriptional nitfr-, as well as borrowed tf, which is likely to be seen in <bwd(y)sdf> bdisaf Bodhisattva and <sdf> saf being (sattva-).21 This approach appears to account for the data in an economic way and motivate the dissimilation in afr particularly well. A development of word-final *w > *tf > -f(t) also seems to be more straightforward than *w > *f (> *-f ?) > -f(t). Pth. *w > tf is also quite parallel to Sogdian and Khwarezmian *w > f (Sims-Williams 2004: 541, 543), agreeing with these being closely related languages (Sims-Williams ibid.), and Bactrian *f (> ) would correspond to the stage of Manich. Pth. word- internal -f-. Alternatively, one could consider an interpretation of both inscriptional <tp> and Manich. <df> as f(Jost Gippert, p.c.), comparing it to Avestan fr- (oblique stem of pitarfather), which is likely to reflect /fr-/,22 and to the development of word-internal *w > Pth. <db>, if this is v as per Sims-Williams (2004: 540). However, the assumption implies that one would need to posit word-final developments of *f > -f, *f > -ft, plus *-tf > -f for loanwords to account for <bwd(y)sdf> and <sdf>, a set that is perhaps not altogether compelling. Hence a development *w > *tf > -f(t) appears to be preferable. 4. There is another piece of evidence which is not compatible with the classical view of the development of *w in Western Iranian. MP nixwr- (Manich. <nyxwr->, Pahlavi <nswbl->) hurry, hasten, incite is obviously a cognate of Pth. nifr-, but ahr four and
Sims-Williams 1983: 140). MP shows marg death, but nothing that would correspond to Pth. <mwrt>. Conversely, marg is not found in Parthian. The MP hapax <zydmrgy> (or <zyrmrgy>, Sundermann 1984: 504) ?-death is not likely to be an error for <zdmrgy> birth-death (Desmond Durkin-Meisterernst, p.c.), MacKenzie (apud Sundermann 1984: 504) considers a connection to Avestan ti- life, Sundermann (ibid.) a reading +<zwd> (fast) or +<zwr> (force). 19 Paikuli inscription 21 d1, 03 (cf. Skjrv 1983/I: 49, II: 79 f.). 20 *-w- > Pth. -tf- also seems to be assumed by Weber (1994: 111 n. 11; his only example <ctfr> four is not attested, though). For word-final position, Lentz (1926: 253) and Huyse (2003: 85 n. 125) assume a development *-w- > -ft (with dialectal variant -f) and interpret this as a metathesis, i.e. both also assume an intermediary stage *-tf. Rastorgueva / Molanova (1981: 172) posit a development *w > *f > f for the word-internal position. One could also consider a dissimilation of the dentals in afrast forty < *atfrast and/or afrdas fourteen < *atfrdas, to which afr < *atfr could have been adjusted; but such an explanation would only account for Parthian, not for the parallel developments in other Ir. languages. 21 Bactrian suggests that <bwd(y)sdf> was borrowed from a form with -tf (Sims-Williams 2004: 544). 22 Cf. Beekes (1988: 73, 86, and 235 s.v. ptar-). I am grateful to Michiel de Vaan for pointing out this reference to me.

ihil forty would lead one to expect MP nihr-.

In view of the discussion above and of the laconic note by Sims-Williams (2004: 540) [nixwa:r-] < [nihwa:r-], one might wonder whether the MP development of *w is not as traditionally assumed either, and could posit the assumption that nixwr- < *nihwr- < *niwr- shows the regular MP result of *w in word-internal position. The reduction seen in ahr and ihil would then need to be due to a specific development here as well, which could have operated at the stage of *hw. A reduction of the consonant cluster would seem particularly likely in the multiple clusters arising in *awatam (cf. Av. aarsatm) > *ahwirhat23 (> via *ihwihl or *ihird ?) > ihil forty, whose -h- would have been transferred also to *ahwr > ahr. In word-final position one would need to assume a reduction *w > *-hw > -h, which would operate in the abstract suffix -h (< *-iya-wa-) and in *pw- bridge > *purh > puhl.24 The adverbial suffix -h would need to have generalised h by paradigmatic levelling from -h.25 This approach implies ad-hoc assumptions for ahr, ihil and -h, but accounts for nixwr-, which is otherwise left without explanation.26 Moreover, a development *w > *hw > xw agrees quite well with other MP sound changes: * yields MP h generally (e.g. pahn wide, broad, mhan home vs. Avestan paana-, maana-, Hbschmann 1895: 203). The sequence *hw < *w merges with old *hw < PIE *s, both resulting in MP xw.27 Also parallel is the development of *fw > MP hw (kahwan old < *kafwan, Bailey 1979: 62b, 64b). But this development needs to be later than the change *hw >xw just discussed, as the hw arising from *fw does not yield xw.28 5. 5.1 The interpretation of Pth. <pwrt> purt bridge and <mwrt> murt death suggested in Section 2. implies that Manich. Pth. <t> and <d> encode two different phonemes also in the position after r. Now there appear to be exceptions exactly in this context: according to Boyce (1975: 17), <t> otherwise encodes t, but rarely also d when in the position after r (an archaic spelling), e.g. wrt- besides wrd- (ward-). This raises the question whether <rt> and <rd> are written indiscriminately and refer to the same pronunciation.29 The data are the following:30 inflectional forms of the verb <wrt-> / <wrd-> wart/d- turn;

Under any assumption (*w > *hw or directly > h), * gives ir here in spite of the neighbouring *w. MP h spring and gh place; throne can be explained as deriving from *-u- (Old Persian gu-, cf. note 11) with generalised from the oblique case (cf. Hbschmann 1895: 195, 203, Brandenstein / Mayrhofer 1964: 121); the same applies to Pth. h and gh. 25 Gauthiot (1918: 67) explains -h as ablative-instrumental *-iya-w of the stem *-iya-wa-. 26 Hennings note (1939: 105) about nixwr- as a developed form of nivr- does not explain anything, and the borrowing from Parthian cautiously considered by Weber (1994: 111 n. 11) needs to assume an unprecented substitution of df by xw. 27 On the possibly monophonematic status of MP <xw>, see Weber 1994. 28 The New Persian merge assumed by Weber (1994: 113) for MP hw and xw (or rather /xw/) is obscure to me; in fact, MP hw yields NP hu (kahun, kuhan old < MP kahwan) while xw gives NP xu (saxun, suxan speech < MP saxwan). 29 This phenomenon needs to be distinguished from cases which show a variation <d> / <> (cf. DurkinMeisterernst 2000: 169ff.). These cases include <bw> / <bwd> bd (past stem of <bw-> baw- be) in a proportion 1:4 (Durkin-Meisterernst 2000: 172), a similar proportion holds for pad to, in (<p> / <pd>). The variation <d> vs. <> is found in instances deriving form Old Ir. t. Conversely, the Pth. result from Old Ir. d is always written <d> (e.g. <kd> kad when, Durkin-Meisterernst 2000: 172 n. 36). The remaining cases of <> are orthographic variants of <t> (Boyce 1975: 17). 30 Corresponding Manich. MP words (where attested) have only <rd>.
24

23

 its derivatives <wrd(g)> prisonerer, <wrdy(y)wn> wagon; its compounds and their derivatives: inflectional forms of <mwrt-> / <mwrd-> amwart/d- collect with <mwrdn> assembly (place), <mwrdyn> collection, <mwrtdnyft> assembly; one inflectional form of <zwrt-> iz-wart/d- return with <zwrdyn> return; one inflectional form of <prwrt-> par-wart/d- prevail31 vs. fra-wart/din <frwrdg> letter (roll); <rt> (< *arta-, Avestan aa-, Old Persian arta, OInd. t-) besides <rdw> (< *artan-, cf. Avestan aauuan-, OP artvan-, OInd. tvan-); both occur only in connection with <prwhr> in a designation of the ether (one of the Manich. elements of light). <rt> could be an archaism of the religious language as is its cognate wrt- /urta-/ in the Sogdian version of the prayer Am voh (Nicholas Sims-Williams, p.c., cf. Gershevitch 1976). If one explains <rt> as an archaism or a borrowing from an older stage of the language, Pth. wart/d- turn with compounds and derivatives is the only case of a variation <d> / <t> in Manich. Pth. orthography.32 At the same time, wart/d- is the only instance of Pth. <rt> other than <pwrt> and <mwrt>.33 The remaining cases are loanwords or unclear: <srt> srt caravan and <s()rtw> sartw caravan leader are borrowed from OInd. srtha- and srthavha- (as is Sogdian s rth, Sims-Williams 1983: 133, 135, 140); two items are unclear: the hapax <wrtdgyft> (thus Sundermanns reading of <()wr(t.gy)ft>, cf. DMD 70a), perhaps it belongs to <wrt/d->; and <mrtyn> (twice attested), for which Henning (apud Sundermann 1973: 115) assumes a connection to Avestan aa-.34 5.2 For the evaluation of the orthography <rt/d>, the following points could be relevant: Old Ir. t usually gives Pth. <d> post-vocalically and after sonorants, and also after r, e.g. <mrd> mard man (Av. marta-), <mrdyft> mardft manliness, <srd> sard cold (Av. sarta-), <srdg> sardg cold (noun), <wxrd> wxard eaten (< *hwar-ta-), <wxrdyg> wxardg meal, <nbrd> nibard battle, <nbrdg> nibardag warlike, <kyrd> kird done (Av. krta-), <kyrdgr> kirdagr mighty, <dyrd> dird held (Av. drta-).35 The voiced counterpart, Old Ir. rd, mostly yields Pth. r, e.g. <zyrd> zir heart (< Proto-Ir. *zdaya-).36 However, Old Ir. ard gives Pth. r (Rastorgueva / Molanova 1981: 162), e.g. <wr> wr flower (Av. vara-), <sr> sr year (Av. sara-). So there is an opposition between -rd < Old Ir. -rt and -r < Old Ir. -rd only for vowels other than a, but no ar < ard vs. ard < art. Connecting the Pth. data to developments in other Ir. languages, one might wonder whether the mixed orthography <rt/d> after a intended to mark a specific pronunciation for

Sims-Williams (1989: 325) connects Pth. <prwrt-> to Sogdian prwrt turn, change, become (< *pari-wart-) and translates the attestation <wd wd tftwdyg | wwd ny prwrtyd> (verse) as (...) and the searing wind does not prevail there. Maybe one could also consider a meaning within the semantic range of the other <()wrt/d->, e.g. and the searing wind does not swirl there or even and the wind does not turn searing there, interpreting <prwrtyd> in the light of its Sogdian cognate. 32 The statement by Boyce quoted in Section 5.1 and the note by Durkin-Meisterernst (2000: 173) to the same effect thus need to be adjusted. 33 Pth. art is also found in names from other languages (Sanskrit, Turkic). 34 Another example might be the unclear hapax <hwwr>, perhaps having good ? (but maybe this is not a complete word, cf. DMD 192a), if <> here is a graphic variant of <t> and not of <d> (cf. note 29). 35 For examples of *t in labial context see Section 2.1. 36 The opposition between voiced stops (from Old Ir. word-internal voiceless stops) and fricatives (from Old Ir. word-internal voiced stops) is not marked in the Manich. script, but has generally been assumed at least for the older stages of Parthian. Sundermann (1989: 123) assumes a merge of both series for Late Middle Parthian (6th c. AD), thus also Rastorgueva / Molanova (1981: 160). See Korn (2010: 424f.) for more discussion.

31

which there was no orthographic conventionperhaps voiceless r + t as Durkin-Meisterernst (2000: 173) assumes. Similarly, Av. <>, which is the result of rt in certain contexts, has been assumed to represent voiceless r, retroflex , or a fricative similar to Czech (Hoffmann 1986: 173ff., de Vaan 2003: 602). Also noteworthy is the occasional lengthening of Av. a preceding <>, e.g. xva- food < *hwar-ta-.37 In Balochi, Old Ir. *art gives rt and *ard gives (e.g. wr-t eats vs. war- otherwise; gwag blossom38 vs. Av. vara-) while rt and rd after other vowels are preserved.39 Pashto likewise has retroflex from Old Ir. rt and rd, but this is independent of the preceding vowel.40 So if the Pth. orthography <rt/d> did indicate a specific sound or sound cluster, the result of *art would arrange itself with similar phenomena in other Ir. languages. It is not clear, though, why a variation <rt/d> is only found with the family <wrt/d-> and not with other words containing Old Ir. *art, or why a specific pronunciation is only marked for wart/d-.41 Perhaps the variation <rt/d> marks the word-internal development, which is exclusively found in the only Pth. present stem with Old Ir. art,42 while the word-final position shows the expected <rd> ard. Inflectional forms and derivatives such as <mrdn> mardn (plural), <mrdyft> mardft, etc., were surely related to <mrd> mard man by the speakers and thus not undergo word-internal development, while a present stem mostly occurs with endings. If <rt/d> is the word-internal development, it is maybe less likely that <rt/d> stands for a devoicing which would not have taken place in word-final position, and a retroflex or fricative output would seem more likely. 6. Summarising the argument above, Manich. Pth. <t> and <d> encode two different phonemes also in the position after r, and that Pth. <pwrt> purt bridge and <mwrt> murt death are to be read as purt and murt. These words are likely to go back to *pw- (the form from which MP puhl also derives) and *mw- (while Sogdian mwrw derives from*muwith generalised ). These are the oblique stems of *ptu- and *mtu-, the former familiar from Av. prtu-, the latter otherwise only found in Sogdian. Pth. <pwrt> bridge and <mwrt> are, then, additional evidence for Sims-Williams claim that Proto-Ir. *w does not yield Parthian f as previously assumed, but results in a consonant group, which would be reduced in Pth. *purft and * murft. By the logic suggested here, -ft would be the Pth. word-final outcome of*w in Manich. Parthian (vs. -f in inscriptional Parthian) vs. -tf- (thus in inscriptional Parthian) > -f- (Manich.) in word-

Cf. de Vaan (2003: 54f., 104, 596). Among the instances relevant here is a- quick; firmament (from the same root as Pth. nifr-), if this does not contain old (de Vaan, ibid.) 38 Thus Sayad Ganj p. 704. Barker / Mengal (1969/II: 463) note gwig wild yellow tulip. 39 Cf. Korn (2005: 97, 189, 220). 40 Cf. Skjrv (1989: 404). A change of r + dental to retroflexes is common cross-linguistically (thus e.g. in Swedish and in Franconian dialects). 41 Sogdian influence cannot be responsible for the orthography of Pth. <wrt/d->: the variation of </d> and <t>, specifically after r, noted by Gershevitch (1954: 42f., 268ff.) does not exist; rather, a late stage of Sogdian probably had [d] as an allophone of /t/ in voiced contexts, thence some cases of C <d> for what is otherwise <t> (Nicholas Sims-Williams, p.c., cf. Sims-Williams 1985: 163 n. 1). Sogdian compounds and derivatives corresponding to Pth. <wrd/t-> are well attested, and always written with <rt>, e.g. prw()rt- turn, zw()rtreturn, wrtn wagon; the interpretation of wrt (Frag. Len. 93, 8) is not clear, but it is unlikely to show *wrturn (Pavel Lurje and Nicholas Sims-Williams, p.c.). Perhaps a denominative verb wr- rain is present here (Yutaka Yoshida, p.c.). 42 Pth. and MP (Pahlavi) nibard- fight are probably denominative formations from nibard battle, cf. the secondary past stems Pth. nibardd (which is the only attested form of the Pth. verb) and MP nibardd (not from the zero grade), cf. OInd. pt.

37

internal position. Middle Persian may likewise show a consonant cluster as the result of *w, yielding *hw > xw. In ahr four and ihil forty, specific processes must then have been at work to effect the simple h; these would be parallel to cluster reductions in these numbers in other Ir. languages. Table 1 presents the Pth. sound changes of r and * + dental discussed in this paper in comparison with some data of selected Western Ir. languages. Table 1. Development of *r, * + dental Proto-Iranian Parthian <yrd> ir <wrd> ur <yrd> ird <wrd> urd <r> r Zazaki Balochi Middle Persian

*d *t

e l ? erd rd ar

ird, urd irt, urt uhl ? rt

il, ul ird, urd uhl

Manich. <r> Pahl. <l>43

*w / [+ lab.]_ <wrt> urt *ard *Hd *art *Ht

word-internal <-rt/d-> -a()- ? erd word final <-rd> -ard

rd

Examples for Zazaki include the cognates of Pth. words mentioned above: for *ard: se year; *d: zei heart (Paul 1998: 169), vlke flower; *t: berd-, kerd- (past stems of do and carry away); *art: serd cold; *w: prd bridge (cf. Section 2.1). Since *t appears to give erd also in labial context (berd- < *bta-), one could perhaps consider vlke a loanword (thus Paul 1998: 169), so that the regular output of *d in labial context could be e or perhaps (cf. e.g. p full, which at least shows * in labial context although not *t). In Balochi, the contexts in which * yields ir and ur are not identical to those of MP and Pth. ir, ur. While ir is the result in palatal contexts and ur in labial ones, the neutral context shows Balochi ur, but MP / Pth. <yr> ir, e.g. Balochi turs- vs. MP, Pth. <tyrs-> be afraid, kurt vs. MP, Pth. <kyrd> done. Other examples include *d: zird heart, *ard: gwag blossom, *art: srt cold. Owing to the absence of other examples for the context *w, it is impossible to decide whether Balochi puhl bridge is a MP loanword or not (Korn 2005: 143148, 328, 121).

Bibliography Bailey, Harold W. 1979. Dictionary of Khotan Saka. Cambridge etc.: Cambridge University Press Barker, Muhammad A. and Aqil Khan Mengal 1969. A Course in Baluchi. Montreal: McGill University Press, 2 vol.

The Pahlavi orthography is ambiguous and could also stand for r (then identical with the Pth. outpout), cf. Hoffmann (1986: 183 n. 38). At any rate, New Persian has l in relevant words.

43

Beekes, Robert S.P. 1988. A Grammar of Gatha-Avestan. Leiden etc.: Brill Benveniste, mile 1940. Textes sogdiens, dits, traduits et comments. (Mission Pelliot 3.) Paris: Paul Geuthner Boyce, Mary 1954. The Manichaean Hymn-Cycles in Parthian. (London Oriental Series 3.) London etc.: Oxford University Press Boyce, Mary 1975. A Reader in Manichaean Middle Persian and Parthian. (Acta Iranica 9.) Leiden etc.: Brill Boyce, Mary 1977. A Word-list of Manichaean Middle Persian and Parthian. (Acta Iranica 9a.) Leiden etc.: Brill Brandenstein, Wilhelm and Manfred Mayrhofer 1964. Handbuch des Altpersischen. Wiesbaden: Harrassowitz DMD = Desmond Durkin-Meisterernst 2004. Dictionary of Manichaean Middle Persian and Parthian. (Dictionary of Manichaean Texts III: Texts from Central Asia and China 1. Corpus Fontum Manichaeorum Subsidia.) Turnhout: Brepols DMG = Desmond Durkin-Meisterernst (no year). Grammatik des Westmitteliranischen

(Parthisch und Mittelpersisch) auf Grund manichischer Texte, des Inschriftenmaterials und auszugsweise der Pahlav-Literatur. Mnster (unpublished manuscript)
Durkin-Meisterernst, Desmond 2000. Erfand Mani die manichische Schrift?, in Ronald E. Emmerick, Werner Sundermann and Peter Zieme (eds), Studia Manichaica. IV. Internationaler Kongre zum Manichismus, Berlin, 14.18. Juli 1997. Berlin: BerlinBrandenburgische Akademie der Wissenschaften, pp. 161178 EWAia = Manfred Mayrhofer 19862001. Etymologisches Wrterbuch des Altindoarischen. Heidelberg: Winter, 3 vol. Gauthiot, Robert 1918. De la rduction de la flexion nominale en iranien, Mmoires de la socit de linguistique de Paris 20 (191618), pp. 6176 Gershevitch, Ilya 1946. Sogdian Compounds, Transactions of the Philological Society 1945, pp. 137149 (= id.: Philologia Iranica [Beitrge zur Iranistik 12]. Wiesbaden: Reichert 1985, pp. 618) Gershevitch, Ilya 1954. A Grammar of Manichean Sogdian. Oxford: Oxford University Press Gershevitch, Ilya 1976. Appendix. In: Sims-Williams 1976, pp. 7582 Gharib 1995 = Badrozzamn Qarb, Sogdian Dictionary: Sogdian Persian English / Farhang-e sod: sod frs engls. Tehran: Farhangan 1374 h.. Ghilain, Antoine 1939. Essai sur la langue parthe, son systme verbal daprs les textes manichens du Turkestan oriental. (Bibliothque du Muson 9.) Louvain: Muson (repr. 1966) Gippert, Jost 2001. Zum eigenen Tod des Kambyses, in Heiner Eichner, Peter-Arnold Mumm, Oswald Panagl and Eberhard Winkler (eds), Fremd und Eigen. Untersuchungen

zu Grammatik und Wortschatz des Uralischen und Indogermanischen in memoriam Hartmut Katz. Vienna: Edition Praesens, pp. 1526 Henning, Walter B. 1937. A list of Middle Persian and Parthian words, Bulletin of the School of Oriental Studies 9, 7992 (= id.: Selected papers I. (Acta Iranica 14.) Leiden
etc.: Brill 1977, 559572) Henning, Walter B. 1939. Sogdian Loan-words in New Persian, Bulletin of the School of Oriental Studies 10, pp. 93106 (= Selected papers I, pp. 639652) Henning, Walter B. 1958. Mitteliranisch, in Handbuch der Orientalistik I, iv, 1, pp. 20130 Hoffmann, Karl 1986. Avestisch , in Rdiger Schmitt and Prods O. Skjrv (eds), Studia grammatica iranica. Festschrift fr Helmut Humbach. (MSS-Beiheft 13.) Munich, pp. 163183 (= id.: Aufstze zur Indorianistik 3. Wiesbaden: Reichert 1992, pp. 837857) Hbschmann, Heinrich 1895. Persische Studien. Strassburg: Trbner Huyse, Philip 2003. Le y final dans les inscriptions moyen-perses et la loi rhytmique proto9

moyen-perse. (Studia Iranica Cahier 29.) Paris: Association pour lavancement des tudes
iraniennes Korn, Agnes 2005. Towards a Historical Grammar of Balochi. Studies in Balochi Historical Phonology and Vocabulary. (Beitrge zur Iranistik 26.) Wiesbaden: Reichert Korn, Agnes 2010. Parthian , Bulletin of the School of Oriental and African Studies 73, pp. 415-436 Lentz, Wolfgang 1926. Die nordiranischen Elemente in der neupersischen Literatursprache bei Firdosi, Zeitschrift fr Indologie und Iranistik 4, pp. 251316 LIV = Helmut Rix et al. 2001. Lexikon der indogermanischen Verben: die Wurzeln und ihre Primrstammbildungen. Wiesbaden: Reichert, 2nd ed. Paul, Ludwig 1998. The Position of Zazaki among West Iranian Languages, in Nicholas Sims-Williams (ed), Proceedings of the Third European Conference of Iranian Studies held in Cambridge, 11th to 15th September 1995. Part I: Old and Middle Iranian Studies. Wiesbaden: Reichert, pp. 163177 Rastorgueva, Vera S. and E. K. Molanova 1981. Parfjanskij jazyk, in Osnovy iranskogo jazykoznanija 2. Moscow, pp. 147232 Sayad Ganj = Sayad Hm 2000. Sayad Ganj: The First Balochi Dictionary / Sayad Gan: Bal awwal baznt balad. Karachi: Sayad Hashmi Academy Schmitt, Rdiger 1989. Compendium Linguarum Iranicarum. Wiesbaden: Reichert Sims-Williams, Nicholas 1976. The Sogdian Fragments of the British Library, Indo-Iranian Journal 18, pp. 4382 Sims-Williams, Nicholas 1983. Indian elements in Parthian and Sogdian, in Klaus Rhrborn and Wolfgang Veenker (eds), Sprachen des Buddhismus in Zentralasien. Vortrge des Hamburger Symposions vom 2. Juli bis 5. Juli 1981. (Verffentlichungen der Societas Uralo-Altaica 16.) Wiesbaden: Harrassowitz, pp. 132141 Sims-Williams, Nicholas 1985. The Christian Sogdian Manuscript C2. (Berliner Turfantexte 12.) Berlin: Akademie-Verlag Sims-Williams, Nicholas 1989. A New Fragment from the Parthian Hymn-cycle Huyadagmn, in Charles-Henri de Fouchcour and Philippe Gignoux (eds), tudes irano-aryennes offertes Gilbert Lazard. (Studia Iranica Cahier 7.) Paris: Association pour l'avancement des tudes iraniennes, pp. 321331 Sims-Williams, Nicholas 2004. The Parthian Abstract Suffix -yft, in John H.W. Penney (ed), Indo-European Perspectives. Studies in Honour of Anna Morpurgo Davies. Oxford New York: Oxford University Press, pp. 539547 Skjrv, Prods O. 1983. The Sassanian Insciption of Paikuli 3. Wiesbaden: Reichert, 2 vol. Skjrv, Prods O. 1989. Pashto, in Schmitt 1989, pp. 384410 Sundermann, Werner 1973. Mittelpersische und parthische kosmogonische und Parabeltexte der Manicher. (Berliner Turfantexte 4.) Berlin: Akademie-Verlag Sundermann, Werner 1982. Die Bedeutung des Parthischen fr die Verbreitung buddhistischer Wrter indischer Herkunft, Altorientalische Forschungen 9, pp. 99-113 (= id.: Manichaica Iranica I. (Serie Orientale Roma 89.) Rome: Istituto italiano per lAfrica e lOriente, pp. 165-181) Sundermann, Werner 1984. Ein weiteres Fragment aus Manis Gigantenbuch, in Orientalia J. Duchesne-Guillemin emerito oblata. (Acta Iranica 23.) Leiden: Brill, pp. 491505 (= id.: Manichaica Iranica II. (Serie Orientale Roma 89.) Rome: Istituto italiano per lAfrica e lOriente, pp. 615-631) Sundermann, Werner 1989. Parthisch, in Schmitt 1989, pp. 114137 Tedesco, Paul 1921. Dialektologie der mitteliranischen Turfantexte, Monde Oriental 15, 184258 de Vaan, Michiel 2003. The Avestan Vowels. (Leiden Studies in Indo-European 12.)
10

Amsterdam New York: Rodopi Weber, Dieter 1994. Zur Problematik eines Phonems /xw/ im Pahlavi, Studia Iranica, Mesopotamica & Anatolica 1, pp. 107118 Yoshida Yutaka 2008: Die buddhistischen sogdischen Texte in der Berliner Turfansammlung und die Herkunft des buddhistischen sogdischen Wortes fr bodhisattva. In: Acta Orientalia Academiae Scientiarum Hungaricae 61, pp. 325-358

11